INDEX

Explanations

programming constructs and code syntax

oai_token-act-pair · gpt-4o-mini Triggered by @bot

code formatting and declarations

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 surla

-0.49

RTLR

-0.48

 mijne

-0.48

 wijze

-0.48

 linkovi

-0.46



-0.46

లాలు

-0.45

 OMITBAD

-0.45

rzost

-0.44

 <=",

-0.42

POSITIVE LOGITS

:✨

0.59

 bala

0.54

IntoConstraints

0.53

 Caravan

0.44

LabelTagHelper

0.43

Viki

0.41

ंदीखरीदारी

0.40

 Governor

0.39

ListItemIcon

0.39

 IFTT

0.39

Activations Density 0.087%

programming constructs and code syntax

code formatting and declarations

No Comments

No Known Activations